Berder wrote:The opinion of anybody has no place in data analysis, unless you're taking an opinion survey.
Which is totally reasonable, if that "data analysis" is a question about winrate. If it's a question like "Is Trog a strong god," though, there are a bunch of
really obvious confounding factors in the winrate data, which is all I see being discussed. Maybe there's some objective set of data that would conclusively show, one way or another, that Trog is a strong god; I expect that said data set would be a lot more comprehensive than raw data about winrates, no matter how you sample the winrate data.
When it comes to evaluating hazy things like game design, opinions do matter. Crawl isn't a machine, it's a game.
I quoted "proved" because you often use that word to refer to statistical analysis
No I don't. Find another time when I used it in that context.
ok.Edited to add: I feel compelled to point out that that's the first hit in Google when you search "site:crawl.develz.org berder proof," fwiw.
than gammafunk applied, you mean. Yes, it was an exaggeration to say that he "proved" the strength of Trog, but his analysis is good enough to convince me.
Oh, right, I see gammafunk talking about how he's "proving" stuff with winrate data all the time.